purr-data issueshttps://git.purrdata.net/jwilkes/purr-data/-/issues2021-01-11T12:39:52Zhttps://git.purrdata.net/jwilkes/purr-data/-/issues/697Gui objects causes freeze/crash2021-01-11T12:39:52ZRanjith HegdeGui objects causes freeze/crashOn latest version of Purr-data from arch OBS (see below)
When I use objects such as MouseState from cyclone, many GUI abstractions, or playlist object from Unauthorized, purr-data either gives a long line of legacy tcl errors (see below...On latest version of Purr-data from arch OBS (see below)
When I use objects such as MouseState from cyclone, many GUI abstractions, or playlist object from Unauthorized, purr-data either gives a long line of legacy tcl errors (see below) or just becomes unresponsive and crashes.
Version -
```
nw.js version 0.28.1
Pd has started the GUI
canvasinfo: v0.1
stable canvasinfo methods: args dir dirty editmode vis
classinfo: v.0.1
stable classinfo methods: size
objectinfo: v.0.1
stable objectinfo methods: class
pdinfo: v.0.1
stable pdinfo methods: dir dsp version
[import] $Revision: 1.2 $
[import] is still in development, the interface could change!
compiled against Pd-l2ork version 2.15.2 (20201102-rev.2d04705d)
```
error (on MouseState)
```
legacy tcl command at 273 of ../shared/hammer/gui.c: proc hammergui_exithook {cmd op} {proc pd {} {}}
legacy tcl command at 274 of ../shared/hammer/gui.c: if {[info tclversion] >= 8.4} {...
legacy tcl command at 277 of ../shared/hammer/gui.c: proc hammergui_ispatcher {cv} {
legacy tcl command at 278 of ../shared/hammer/gui.c: if {[string range $cv 0 1] == ".x"
legacy tcl command at 279 of ../shared/hammer/gui.c: && [string range $cv end-1 end] == ".c"} {
legacy tcl command at 280 of ../shared/hammer/gui.c: return 1} else {return 0}
legacy tcl command at 281 of ../shared/hammer/gui.c: }
legacy tcl command at 283 of ../shared/hammer/gui.c: proc hammergui_remouse {} {
legacy tcl command at 284 of ../shared/hammer/gui.c: bind all <<hammerdown>> {}
legacy tcl command at 285 of ../shared/hammer/gui.c: bind all <<hammerup>> {}
legacy tcl command at 286 of ../shared/hammer/gui.c: pd [concat #hammergui _remouse \;]
legacy tcl command at 287 of ../shared/hammer/gui.c: }
legacy tcl command at 289 of ../shared/hammer/gui.c: proc hammergui_mousexy {target} {
legacy tcl command at 290 of ../shared/hammer/gui.c: set x [winfo pointerx .]
legacy tcl command at 291 of ../shared/hammer/gui.c: set y [winfo pointery .]
legacy tcl command at 292 of ../shared/hammer/gui.c: pd [concat #hammermouse $target $x $y \;]
legacy tcl command at 293 of ../shared/hammer/gui.c: }
legacy tcl command at 296 of ../shared/hammer/gui.c: global hammergui_ispolling
legacy tcl command at 297 of ../shared/hammer/gui.c: global hammergui_x
legacy tcl command at 298 of ../shared/hammer/gui.c: global hammergui_y
legacy tcl command at 299 of ../shared/hammer/gui.c: set hammergui_ispolling 0
legacy tcl command at 300 of ../shared/hammer/gui.c: set hammergui_x 0
legacy tcl command at 301 of ../shared/hammer/gui.c: set hammergui_y 0
legacy tcl command at 303 of ../shared/hammer/gui.c: proc hammergui_poll {} {
legacy tcl command at 304 of ../shared/hammer/gui.c: global hammergui_ispolling
legacy tcl command at 305 of ../shared/hammer/gui.c: global hammergui_x
legacy tcl command at 306 of ../shared/hammer/gui.c: global hammergui_y
legacy tcl command at 307 of ../shared/hammer/gui.c: if {$hammergui_ispolling == 1} {
legacy tcl command at 308 of ../shared/hammer/gui.c: set x [winfo pointerx .]
legacy tcl command at 309 of ../shared/hammer/gui.c: set y [winfo pointery .]
legacy tcl command at 310 of ../shared/hammer/gui.c: if {$hammergui_x != $x || $hammergui_y != $y} {
legacy tcl command at 311 of ../shared/hammer/gui.c: pd [concat #hammermouse _poll $x $y \;]
legacy tcl command at 312 of ../shared/hammer/gui.c: set hammergui_x $x
legacy tcl command at 313 of ../shared/hammer/gui.c: set hammergui_y $y
legacy tcl command at 314 of ../shared/hammer/gui.c: }
legacy tcl command at 315 of ../shared/hammer/gui.c: after 50 hammergui_poll
legacy tcl command at 316 of ../shared/hammer/gui.c: }
legacy tcl command at 317 of ../shared/hammer/gui.c: }
legacy tcl command at 319 of ../shared/hammer/gui.c: proc hammergui_refocus {} {
legacy tcl command at 320 of ../shared/hammer/gui.c: bind Canvas <<hammerfocusin>> {}
legacy tcl command at 321 of ../shared/hammer/gui.c: bind Canvas <<hammerfocusout>> {}
legacy tcl command at 322 of ../shared/hammer/gui.c: pd [concat #hammergui _refocus \;]
legacy tcl command at 323 of ../shared/hammer/gui.c: }
legacy tcl command at 325 of ../shared/hammer/gui.c: proc hammergui_revised {} {
legacy tcl command at 326 of ../shared/hammer/gui.c: bind Canvas <<hammervised>> {}
legacy tcl command at 327 of ../shared/hammer/gui.c: bind Canvas <<hammerunvised>> {}
legacy tcl command at 328 of ../shared/hammer/gui.c: pd [concat #hammergui _revised \;]
legacy tcl command at 329 of ../shared/hammer/gui.c: }
legacy tcl command at 362 of ../shared/hammer/gui.c: event add <<hammerdown>> <ButtonPress>
legacy tcl command at 363 of ../shared/hammer/gui.c: event add <<hammerup>> <ButtonRelease>
legacy tcl command at 142 of ../shared/hammer/gui.c: bind all <<hammerdown>> {pd [concat #hammergui _up 0 \;]}
legacy tcl command at 144 of ../shared/hammer/gui.c: bind all <<hammerup>> {pd [concat #hammergui _up 1 \;]}
```https://git.purrdata.net/jwilkes/purr-data/-/issues/696[webapp] semicolon escapes with a backslash2020-12-03T21:53:25ZZack Lee[webapp] semicolon escapes with a backslashIn Purr Data webapp, typing "; pd dsp 1" into a message box escapes the ; with an unwanted \ backslash.In Purr Data webapp, typing "; pd dsp 1" into a message box escapes the ; with an unwanted \ backslash.https://git.purrdata.net/jwilkes/purr-data/-/issues/695ofelia library cannot load in Purr Data2020-12-23T22:00:02Z60-hzofelia library cannot load in Purr DataOfelia library cannot be loaded with [declare -lib ofelia] or [import ofelia]
I guess because it is build against Pd-0.51-0 and pd-l2ork is still pd-0.48.0
Would it be possible to use this wonderful library with purr data one day? It is...Ofelia library cannot be loaded with [declare -lib ofelia] or [import ofelia]
I guess because it is build against Pd-0.51-0 and pd-l2ork is still pd-0.48.0
Would it be possible to use this wonderful library with purr data one day? It is more modern, stable and platform consistent than Gem.https://git.purrdata.net/jwilkes/purr-data/-/issues/694crash after segfault2020-11-11T23:16:22ZDave Riedstracrash after segfaultpurr-data has segfaulted on me a few times over the past few days, so I thought I'd open an issue to collect information. I don't think this is bad enough to warrant anyone working on it just yet, but if other people notice this behaviou...purr-data has segfaulted on me a few times over the past few days, so I thought I'd open an issue to collect information. I don't think this is bad enough to warrant anyone working on it just yet, but if other people notice this behaviour maybe they should +1.
This seems similar to #688 (same frequency once every 2-3 hours, unpredictable, also doesn't seem to be tied to DSP) but the output is different:
```
$ purr-data
guidir is /opt/purr-data/lib/pd-l2ork/bin
priority 6 scheduling enabled.
[91426:91453:1105/114452.115529:ERROR:nss_util.cc(724)] After loading Root Certs, loaded==false: NSS error code: -8018
[91461:91461:1105/114452.211665:ERROR:sandbox_linux.cc(377)] InitializeSandbox() called with multiple threads in process gpu-process.
priority 8 scheduling enabled.
priority 8 scheduling enabled.
[91461:91461:1105/114844.312594:ERROR:buffer_manager.cc(453)] [.DisplayCompositor-0x55592e447360]GL ERROR :GL_INVALID_OPERATION : glBufferData: <- error from previous GL command
[91426:91426:1105/121046.928048:ERROR:extension_function_dispatcher.cc(601)] Permission denied for nw.App.quit
Segmentation fault (core dumped)
```
journal:
```
Nov 05 16:18:57 gedachte kernel: purr-data[111872]: segfault at 561e8dc13b30 ip 0000561e8dc13b30 sp 00007fffed821448 error 15
Nov 05 16:18:57 gedachte kernel: Code: 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 <00> 00 00 00 00 00 00 00 91 01 00 00 00 00 00 00 70 84 d0 8d 1e 56
```
per discussion on the other issue I've removed the Gem library and recently faustgen2~ as well. also, I think each time purr-data crashes it's when I'm deleting a number box.
system info:
* purr-data: `Pd-L2Ork-2.15.2 (20201024-rev.c3462eac)`
* OS: `Pop!_OS 20.10` with Gnome 3.38.1 desktop on Wayland
* jackd: `1.9.12`https://git.purrdata.net/jwilkes/purr-data/-/issues/693Changing font in main patch messes up GOP rendering2020-10-30T22:32:19ZAlbert GräfChanging font in main patch messes up GOP renderingThis is a regression in 2.14.0 and later, 2.13.0 doesn't exhibit this misbehaviour. The following screen capture is from 2.14.0:
![Peek_2020-10-30_23-25](/uploads/08fe0d724a75e1aaab330b9d3313847e/Peek_2020-10-30_23-25.gif)
Note the art...This is a regression in 2.14.0 and later, 2.13.0 doesn't exhibit this misbehaviour. The following screen capture is from 2.14.0:
![Peek_2020-10-30_23-25](/uploads/08fe0d724a75e1aaab330b9d3313847e/Peek_2020-10-30_23-25.gif)
Note the artifact at the end of the gif animation (2nd "ghost" gatom showing in gop where there is actually only one).https://git.purrdata.net/jwilkes/purr-data/-/issues/691Needed: Automatic detection of dependencies on Windows2020-10-07T08:13:38ZAlbert GräfNeeded: Automatic detection of dependencies on WindowsAs #690 shows again, the dependencies of some externals we ship on Windows keep changing. A proper solution is in order, along the lines of the dependencies script that we have on the Mac, which takes care of at least some of the most im...As #690 shows again, the dependencies of some externals we ship on Windows keep changing. A proper solution is in order, along the lines of the dependencies script that we have on the Mac, which takes care of at least some of the most important and problematic externals, in particular Gem and fluid~. (I think that a full-blown dependencies script, like the one we use on the Mac, would be overkill for now, and in any case it would probably be quite slow on the Windows file system.)https://git.purrdata.net/jwilkes/purr-data/-/issues/689CSS: Text color cannot be distinguished between object, message, comment etc...2020-10-05T10:24:59Z60-hzCSS: Text color cannot be distinguished between object, message, comment etc...I am trying to make a simple readable GUI theme for my workshop using css but ".box_text" field line 270 is binded to all boxes (object, message, comment etc...) which is reducing the design a lot.
`/* text inside boxes: message boxes, ...I am trying to make a simple readable GUI theme for my workshop using css but ".box_text" field line 270 is binded to all boxes (object, message, comment etc...) which is reducing the design a lot.
`/* text inside boxes: message boxes, object boxes, graphs, comments, etc. */`
So this kind of theme which avoid a lot of confusion between message and object boxes is not currently possible because it uses white text for boxes and black text for messages and comment:
![Capture_d_écran_2020-10-05_à_12.20.42](/uploads/1101635cbee00ade43096567d0503d91/Capture_d_écran_2020-10-05_à_12.20.42.png)https://git.purrdata.net/jwilkes/purr-data/-/issues/686text object, message, and comment, when edited are reinstantiated and thrown ...2020-09-23T03:41:39ZIvica Bukvictext object, message, and comment, when edited are reinstantiated and thrown on top of the glist stacktext object, message, and comment, when edited, are reinstantiated and thrown on top of the glist stack which messes with the preset hub instances that rely on externally saved presets and which rely on prerecorded node locations. This r...text object, message, and comment, when edited, are reinstantiated and thrown on top of the glist stack which messes with the preset hub instances that rely on externally saved presets and which rely on prerecorded node locations. This renders saved presets useless. May also need to warn users that to front/to back commands may exacerbate this problem for all objects. May need to be addressed through both retaining the existing object position and improving the hub logic.https://git.purrdata.net/jwilkes/purr-data/-/issues/684[IMPORTANT] Calling sys_close_audio() freezes the app on macOS2020-09-23T18:29:18ZZack Lee[IMPORTANT] Calling sys_close_audio() freezes the app on macOSIn macOS (Catalina 10.15.5), toggling the DSP on and off freezes the Purr Data app. (the latest master branch)
I believe this issue happens from one of the latest releases. (Probably during the GSoC period)
I tried to fix this and it t...In macOS (Catalina 10.15.5), toggling the DSP on and off freezes the Purr Data app. (the latest master branch)
I believe this issue happens from one of the latest releases. (Probably during the GSoC period)
I tried to fix this and it turned out that calling `sys_close_audio()` causes the freezing.
I believe the function itself (and Port Audio's `pa_close_audio()`) is not what's causing it since it has not been updated for a long time.
I could not really fix this issue but found a temporary solution to at least make it work on macOS which is the following:
```
void sys_close_audio(void)
{
#ifdef __APPLE__ /* temporary solution to fix freeze on macOS */
return;
#endif
...
}
```
This will at least allow macOS users to be able to use Purr Data. I tested this and it worked fine.
I think this issue needs to be fixed ASAP because macOS users won't be able to use the app until it is fixed.
I also have one question: Is "use callback" in Audio Preferences set to true by default? I tried unchecking it but it causes the app to either freeze or crash.https://git.purrdata.net/jwilkes/purr-data/-/issues/683Purr Data / pd-l2ork name inconsistencies between systems2020-09-12T16:42:57Z60-hzPurr Data / pd-l2ork name inconsistencies between systems- On OSX the app name is "pd-l2ork.app", the icon is [pd]
- On Linux the App name is "Purr Data", and the icon is [l2Ork]
- On Windows the app name is "Purr Data", (the app is in "Purr Data" with pd.exe executable)
Also the informatio...- On OSX the app name is "pd-l2ork.app", the icon is [pd]
- On Linux the App name is "Purr Data", and the icon is [l2Ork]
- On Windows the app name is "Purr Data", (the app is in "Purr Data" with pd.exe executable)
Also the information panel say in every systems: "About Pd-l2ork"
A better consistency could be possible and helpful for newcomers?https://git.purrdata.net/jwilkes/purr-data/-/issues/682Auto-complete feature in Purr Data2020-09-11T21:28:56Z60-hzAuto-complete feature in Purr DataPurr Data still miss one great feature which is auto-complete for objects boxes.Purr Data still miss one great feature which is auto-complete for objects boxes.https://git.purrdata.net/jwilkes/purr-data/-/issues/681Fallback mechanism for missing translations2020-09-03T22:54:02ZAlbert GräfFallback mechanism for missing translationsWe need a fallback mechanism for missing translations in foreign languages which defaults to the en locale if some localized string cannot be found. Cf. https://git.purrdata.net/jwilkes/purr-data/-/merge_requests/500#note_7838.We need a fallback mechanism for missing translations in foreign languages which defaults to the en locale if some localized string cannot be found. Cf. https://git.purrdata.net/jwilkes/purr-data/-/merge_requests/500#note_7838.https://git.purrdata.net/jwilkes/purr-data/-/issues/680pddp link does not always revert to its instantiated format after editing2020-09-02T19:38:02ZIvica Bukvicpddp link does not always revert to its instantiated format after editingIn the about patch, in edit mode, click on a pddp/link, and it will be activated and become editable. If one clicks outside any other object, it will be properly deselected and re-instantiate in its clickable (non-activated) format. Howe...In the about patch, in edit mode, click on a pddp/link, and it will be activated and become editable. If one clicks outside any other object, it will be properly deselected and re-instantiate in its clickable (non-activated) format. However, if while pddp/link is activated, if you click on something else and immediately select it, it will become deselected but its content will remain as if it is in activated mode.https://git.purrdata.net/jwilkes/purr-data/-/issues/678Data dialog issues2020-08-28T04:37:25ZIvica BukvicData dialog issuesData dialog should auto-update when the data structure is manipulated directly. This includes vars and the vector.
The second issue is not entirely clear to me. When editing vector data, are we allowed to add new data or only edit exist...Data dialog should auto-update when the data structure is manipulated directly. This includes vars and the vector.
The second issue is not entirely clear to me. When editing vector data, are we allowed to add new data or only edit existing? If we should be able to add new data, then this is impossible since enter is mapped to the dialog's OK button and this should be fixed. A common way of addressing this is (e.g. via social media platforms) when trying to send a multiline comment, one presses shift+return(or shift+enter) which gives them a newline without sending the message.https://git.purrdata.net/jwilkes/purr-data/-/issues/677Adjusting font size back and forth at a zoom level than default (100%) distor...2020-08-27T05:04:25ZIvica BukvicAdjusting font size back and forth at a zoom level than default (100%) distorts location of objects in gop-enabled subpatches and toplevelsResized GOP windows suddenly lose visibility of its contents (which is expected if it is a subpatch whose font is tied to the parent patch). However, resetting font size does not restore original position. Sometimes objects remain stale ...Resized GOP windows suddenly lose visibility of its contents (which is expected if it is a subpatch whose font is tied to the parent patch). However, resetting font size does not restore original position. Sometimes objects remain stale outside the GOP window and one cannot get rid of them.https://git.purrdata.net/jwilkes/purr-data/-/issues/676Subpatch data scalar toplevel window should rescale its scalars when the wind...2020-08-27T05:02:23ZIvica BukvicSubpatch data scalar toplevel window should rescale its scalars when the window is rescaledOpen disis_wiimote-help.pd and the subpatch for either blobs or joystick and resize it and note that the data scalar does note rescale accordingly.Open disis_wiimote-help.pd and the subpatch for either blobs or joystick and resize it and note that the data scalar does note rescale accordingly.https://git.purrdata.net/jwilkes/purr-data/-/issues/675Selection boxes of data scalars in the GOP-enabled subpatch window are way of...2020-08-27T05:01:23ZIvica BukvicSelection boxes of data scalars in the GOP-enabled subpatch window are way off and correlate to the distance from the top-left corner of the patch windowOpen disis_wiimote-help.pd and open the subpatch with blobs, go to edit mode and try selecting and moving the 4 blobs. This may be related to #647Open disis_wiimote-help.pd and open the subpatch with blobs, go to edit mode and try selecting and moving the 4 blobs. This may be related to #647https://git.purrdata.net/jwilkes/purr-data/-/issues/674Clicking on the resize text object that is not a GOP still offers a diagonal ...2021-03-29T14:11:19ZIvica BukvicClicking on the resize text object that is not a GOP still offers a diagonal resize in the bottom right corner that results in an error: canvas: no method for '_click_for_resizing'https://git.purrdata.net/jwilkes/purr-data/-/issues/670key missing in [key] and [keyname]2020-08-19T05:05:15Z60-hzkey missing in [key] and [keyname]The following key reports have problems with [keyname] object (same with [key]):
Space: nothing return
Return: only keydown/keyup but no name with [keyname]
cmd: nothing return
Caps_lock: nothing return
Shift_L: only "Shift" (no info Sh...The following key reports have problems with [keyname] object (same with [key]):
Space: nothing return
Return: only keydown/keyup but no name with [keyname]
cmd: nothing return
Caps_lock: nothing return
Shift_L: only "Shift" (no info Shift_L / Shift_R)
Super_L: nothing returnhttps://git.purrdata.net/jwilkes/purr-data/-/issues/669nw.js 0.46/0.47 issues2020-09-06T08:11:55ZAlbert Gräfnw.js 0.46/0.47 issuesDocument regressions with respect to the latest nw.js versions here.Document regressions with respect to the latest nw.js versions here.